股票市场的非线性结构与混沌效应检验--基于BDS与CR方法  

Testing Nonlinear Structure and Chaos Effect of Chinese Stock Markets Based on BDS and CR Methods

作　　者：马超群[1] 邹琳[1] 李红权[2] 

机构地区：[1]湖南大学工商管理学院,湖南长沙410082 [2]湖南师范大学商学院,湖南长沙410081

出　　处：《湖南大学学报（自然科学版）》2008年第5期85-88,共4页Journal of Hunan University:Natural Sciences

基　　金：国家自然科学基金资助项目(70471030);国家社会科学基金资助项目(030JBY56);教育部博士点基金资助项目

摘　　要：在传统研究方法的基础上,运用Rosenstein提出的小数据量算法计算最大李雅普诺夫指数,进而引入BDS与返回临近检验(CR),从不同角度对中国股票市场的混沌动力学结构进行分析.为了避免破坏混沌吸引子的分形结构,采用对数线性趋势消除法(LLD)进行数据处理.研究结果表明,中国股市具有低维混沌吸引子、对初值敏感依赖性、准周期性等显著的非线性混沌特征.并就市场混沌的经济含义与应用价值进行了探讨.Based on the analysis of the traditional methods, the small data algorithm originated by Rosenstein was introduced to calculate the maximum Lyapunov exponent, and two types of tests-BDS and close return （CR） were used to further analyze the dynamical characteristic of chaotic of Chinese stock markets. For avoiding destroying the fractal structure, we utilized a special method of log-linear detrended （LLD） to process the sample data. The conclusions indicate that the notable chaos dynamics characteristic appear to exhibit in Chinese stock markets, such as a low dimensionality chaos attractor, sensitive dependence on initial values, quasi-periodicity. And the economical implication and application value of chaos was also investigated.

关 键 词：混沌理论 非线性分析 对数线性趋势消除法 BDS检验 返回临近检验 

分 类 号：F830.9[经济管理—金融学]